Visual parking lot call remote service method and system

Abstract

The invention discloses a visual parking lot call remote service method and system, and relates to a parking lot remote call service technology, and the method comprises the following steps: based on the response state of a parking lot entrance gate, combining video data and parking space state change conditions, associating a parking lot synchronous entering vehicle, and carrying out the marking, recording in a parking lot real-time database; and according to the change condition of the parking space state, behavior logic analysis is performed on the vehicle causing the change of the parking space state, a current vehicle abnormal type list analysis report is generated according to a parking lot vehicle call remote service abnormal type list in combination with the gate payment state, and an abnormal judgment result is marked in the abnormal type list analysis report. According to the invention, the abnormal type of the remote call service vehicle of the parking lot is analyzed, the reasonable degree of the abnormal type is judged in combination with the call content of the user, the actual remote service call logic is optimized, and the communication efficiency between an administrator and the user is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of remote call services for parking lots. Specifically, it relates to a method and system for visualizing remote call services in a parking lot. Background Art

[0002] A visual parking lot is an intelligent parking lot that integrates advanced information technologies. It uses various technical means to monitor and display information such as parking spaces and vehicles in the parking lot in real time, thereby achieving efficient management and convenient use of the parking lot, including collecting vehicle video data in the parking lot through cameras and obtaining the real-time status of parking spaces through geomagnetic sensors;

[0003] However, in the actual use process of the parking lot, when a user is leaving the parking lot, if there are difficulties or machine abnormalities at the exit of the parking lot, it is often necessary to contact the manual staff and call for remote services to solve the problems that occur. Due to the information deviation between the user and the parking lot administrator, it is easy to cause inefficient communication and lead to the problem of parking lot congestion, and it is difficult to achieve two-way interaction between the platform and the user, resulting in problems of low practicality and functionality.

[0004] In response to the problems in the related technologies, no effective solutions have been proposed yet. Summary of the Invention

[0005] In response to the problems in the related technologies, the present invention proposes a method and system for visualizing remote call services in a parking lot to overcome the above-mentioned technical problems existing in the existing related technologies.

[0006] To this end, the specific technical solutions adopted by the present invention are as follows:

[0007] A method for visualizing remote call services in a parking lot, the method comprising the following steps:

[0008] S1. Based on the reaction status of the entrance gate of the parking lot, combined with the video data and the change of the parking space status, associate the vehicles entering the parking lot synchronously and make a mark, and record it in the real-time database of the parking lot;

[0009] S2. According to the change of the parking space status, perform behavioral logic analysis on the vehicles that cause the change of the parking space status, combined with the payment status of the gate, and generate an analysis report of the current vehicle abnormal type list according to the list of abnormal types of remote call services for vehicles in the parking lot, and indicate the abnormal determination result in the analysis report of the abnormal type list, including true abnormality and pending abnormality;

[0010] S3. For the user of the currently parked vehicle calling for remote service content, combine the user's call content with the keyword libraries of different abnormal types, determine the consistency between the current user's call content and the abnormal types in the analysis report of the abnormal type list, and mark the rationality. Based on the determination results of different abnormal types, output actual solutions, including gate release and manual transfer. At the same time, present the analysis report of the current vehicle abnormal type list after marking to the administrator's device to systematically provide the basis for the determination result of the user's staying reason. For the vehicles staying at the exit exceeding the restricted time, actively contact the administrator to assist in remote service;

[0011] S4. According to the sensor information of the exit gate, for the vehicles that have left the parking lot, make a note in the form of a numbered file in the real-time database of the parking lot.

[0012] As a preferred implementation manner, S1 includes the following steps:

[0013] S11. Through the geomagnetic sensors deployed in the parking spaces, as well as the exit gate and entrance gate of the parking lot, collect the parking space status and gate data, including the parking space occupancy status, parking space occupancy time, vehicle entry time, license plate recognition result, gate lifting status, and gate payment status;

[0014] S12. Through the gate cameras and parking space cameras deployed in the parking lot, obtain the real-time vehicle video data in the parking lot and extract vehicle features, including identifying the vehicle color through the color histogram and identifying the vehicle contour through the edge detection algorithm. Combine the parking space occupancy status, vehicle entry time, license plate recognition result, and gate lifting status to associate the vehicles entering the parking lot synchronously, including the following steps:

[0015] According to the vehicle color and contour information collected at the entrance gate, based on the parking space occupancy status within n minutes after the vehicle entry time, extract the vehicle color and contour information of the occupied parking space within n minutes when the parking space is occupied. Determine the similarity between the vehicle feature data of the parking space and the vehicle feature data of the entrance gate through the Euclidean distance. Based on the similarity threshold, associate the entry time, license plate recognition result, and parking space number of the current vehicle, and generate a numbered file. Establish a real-time database for the parking lot through MySQL, use the numbered file as the file name, and store relevant information, including the license plate recognition result, parking space number, vehicle entry time, and the results of the current vehicle edge detection recognition and color histogram reading.

[0016] As a preferred implementation manner, S2 includes the following steps:

[0017] S21. According to the change result of the parking space occupancy status in the parking lot, for the vehicles that cause the change in the parking space occupancy status, combine the parking space occupancy time to conduct behavioral logic analysis to determine the departing vehicles;

[0018] S22. For the determined departing vehicle, according to the preset list of abnormal types of parking lot vehicle call remote services, combined with the gate payment status and the video data of the exit camera, generate an analysis report of the current vehicle abnormal type list, and indicate the abnormal relevance and preliminary determination results of different abnormal types in the analysis report of the abnormal type list.

[0019] As a preferred embodiment, S21 includes the following steps:

[0020] S211. According to the change result of the parking space occupancy status in the parking lot, record the time node when the parking space status changes from idle to occupied, and obtain the parking space activation timestamp t start And obtain the license plate recognition result according to the mark number associated with the current parking space number;

[0021] S212. Record the time node when the parking space status changes from occupied to idle, and obtain the parking space deactivation timestamp t over , calculate and obtain the vehicle usage time Δt of the current parking space = t over -t start , determine whether the vehicle leaving the current parking space is a departing vehicle. When Δt > t θ , determine that the vehicle in the current parking space is a departing vehicle. When Δt ≤ t θ , determine that the current vehicle is a pending vehicle, where t θ is the parking space usage time threshold;

[0022] S213. For the pending vehicle, make a secondary determination in combination with the remaining idle parking space status in the parking lot to determine the actual status of the current pending vehicle. The steps include: extracting the parking space related data where the parking space status changes from idle to occupied within n / 2 minutes. When there are parking spaces without generated mark numbers, extract the vehicle color and contour information of all parked vehicles in the parking spaces without generated mark numbers, and perform Euclidean distance similarity determination with the vehicle edge detection recognition and color histogram reading results of the current pending vehicle in the parking lot real-time database, and update the parking space information that meets the similarity determination result to the current pending vehicle mark number file in the parking lot real-time database;

[0023] When there are no parking spaces without generated mark numbers, determine the current vehicle as a departing vehicle and make a note in the current vehicle file in the parking lot real-time database.

[0024] As a preferred embodiment, S22 includes the following steps:

[0025] S221. Preset the list of abnormal types of parking lot vehicle call remote services, establish an abnormal set E = {E1, E2, E3, E4}, including E1 payment anomaly, E2 license plate entry anomaly, E3 gate lifting anomaly, E4 other anomalies. The determination feature of each anomaly is that when the gate payment information Fpf = 0, license plate information F lp = 1 indicates abnormal payment; when the turnstile payment information F pf = 1, license plate information F lp = 0 indicates abnormal license plate entry; when the turnstile payment information F pf = 1, license plate information F lp = 1 to detect the status of the exit turnstile lifting bar F up , when F up = 1 represents other abnormalities, when F up = 0 indicates abnormal turnstile lifting;

[0026] S222. Extract the data saved in the mark number file of the departing vehicle in the parking lot real-time database, and combine the turnstile payment status and the exit camera video data to generate the list data of abnormal types of the remote service for calling parking lot vehicles for the vehicles staying at the current exit turnstile:

[0027] Identify the license plate recognition result of the vehicle staying in front of the current exit turnstile through the exit turnstile, and traverse the license plate recognition results in the mark number file in the parking lot real-time database. When there is a consistent license plate, determine the license plate information F of the vehicle staying in front of the current exit turnstile lp = 1, when there is no consistent license plate, determine the license plate information F of the vehicle staying in front of the current exit turnstile lp = 0;

[0028] For the case where the license plate information F lp = 1, obtain the turnstile payment information of the current license plate vehicle through the exit turnstile. The payment status is F pf = 1 for paid, F pf = 0 for unpaid;

[0029] For the case where the license plate information F lp = 0, retrieve all the actually received payment amounts M up to now 实收 and the payment amount M of the departing vehicle 出库 , calculate the amount anomaly value ΔM = M 实收 - M 出库 , when ΔM = 0, determine that the payment has been made F lp = 1, when ΔM < 0, determine that the payment has not been made F pf = 0;

[0030] When the exit turnstile completes the lifting, record F up = 1, when the exit turnstile does not complete the lifting, record F up = 0;

[0031] S223. For the abnormal entry of E2 license plates and the abnormal raising of the E3 gate, obtain the video data of the vehicle staying at the current exit gate through the video data of the exit camera. Extract the features of the currently staying vehicle through the color histogram and edge detection algorithms, and determine the Euclidean distance similarity by combining the vehicle features in the current departing vehicle mark number file. When there is a matching mark number file, it is determined that the abnormal entry of the current E2 license plate and the abnormal raising of the E3 gate are true abnormalities. When there is no matching mark number file, the abnormality is determined to be a pending abnormality;

[0032] S224. Based on the F lp 、F pf 、F up of the current staying vehicle, generate an analysis report on the list of abnormal types of the current vehicle. The analysis report includes the abnormal type E x of the current staying vehicle, where the value range of x is 1 - 4. When x = 2 or x = 3, synchronously mark the abnormality as a true abnormality or a pending abnormality.

[0033] As a preferred embodiment, the S3 includes the following sub - steps:

[0034] S31. Transmit the analysis report on the list of abnormal types of the current vehicle to the administrator device, record the content of the current vehicle's remote service call, extract the keywords in the content of the current vehicle's remote service call through natural language processing technology, and determine the abnormality rationality by combining the abnormal types and keywords in the analysis report on the list of abnormal types of the current vehicle;

[0035] S32. Based on the result of the abnormality rationality determination, take targeted measures for the user's remote service call for service.

[0036] As a preferred embodiment, the S31 includes the following steps:

[0037] S311. Preset the keyword set k corresponding to each abnormal type E where k represents the abnormal type, N represents the keyword number, and assign weights to each keyword Convert the content of the current vehicle's remote service call from speech to text through Speech - to - Text, perform word segmentation on the text through the forward maximum matching method, remove stop words and retain the core keywords, and output the word segmentation set G = {g1, g2,..., g M}, and based on the preset synonym table, replace the words in the word segmentation set with standard keywords;

[0038] S312. For each abnormal type E k, count the number and weight of keywords matching the user's text, convert the matching weight into a percentage form, and select the abnormal type with the highest percentage as the target abnormal type in the current vehicle remote service call content;

[0039] S313. Count the abnormal types in the abnormal type list analysis report of the current vehicle. When the abnormal type in the abnormal type list analysis report of the current vehicle is the same as the target abnormal type, it means that the current vehicle's abnormality is reasonable. When the abnormal type is inconsistent with the target abnormal type, it means that the current vehicle's abnormality is unreasonable, and record it in the abnormal type list analysis report of the current vehicle to obtain the abnormal type list rationality analysis report.

[0040] As a preferred embodiment, S32 includes the following steps:

[0041] S321. When the user abnormal type is E2 or E3 and both the true abnormality and the abnormality rationality determination are satisfied, directly perform the operation of releasing the turnstile.

[0042] S322. When the user abnormal type does not meet the requirements of S321, output the abnormal type list rationality analysis report of the current vehicle to the administrator device and perform the manual transfer operation.

[0043] As a preferred embodiment, S322 includes the following steps:

[0044] S3221. Identify the abnormal vehicle at the exit through the camera at the exit turnstile based on the edge detection algorithm, record the start timestamp of the abnormal vehicle occupying the exit, and obtain the cumulative duration t of the abnormal vehicle occupying the exit 累计 , when t 累计 exceeds 5 minutes and the administrator device has not received the user's call for remote service request, actively transmit the abnormal type list rationality analysis report of the current vehicle to the administrator device;

[0045] S3221. Connect the communication device deployed at the exit turnstile through the communication device at the administrator device to actively provide services for the current abnormal vehicle.

[0046] A visual parking lot call remote service system includes a vehicle data collection module, a vehicle behavior analysis module, an abnormal list generation module, a call content determination module, and a real-time operation module;

[0047] The vehicle data collection module collects the parking space status and turnstile data through the geomagnetic sensors deployed in the parking spaces and the exit turnstiles and entrance turnstiles of the parking lot, and collects the vehicle video picture data through the cameras deployed inside the parking lot;

[0048] The vehicle behavior analysis module associates the vehicles entering the parking lot synchronously based on the response status of the entrance gate of the parking lot, combines video data and the changes in the parking space status, makes a mark, and records it in the real-time database of the parking lot;

[0049] The abnormal list generation module conducts a behavioral logic analysis on the vehicles that cause changes in the parking space status according to the changes in the parking space status, combines the payment status of the gate, and generates an analysis report on the abnormal type list of the current vehicle according to the abnormal type list of the remote service for vehicle calls in the parking lot, and indicates the abnormal determination result in the analysis report on the abnormal type list, including true abnormality and pending abnormality;

[0050] The call content determination module conducts a consistency determination on the content of the user's call for remote service of the currently parked vehicle, combines the user's call content with the keyword library of different abnormal types, and determines the consistency between the current user's call content and the abnormal types in the analysis report on the abnormal type list, and marks the rationality;

[0051] The real-time operation module outputs actual solutions based on the determination results of different abnormal types, including gate release and manual transfer, and at the same time presents the analysis report on the abnormal type list of the current vehicle after marking to the administrator's device, so as to systematically provide the basis for the determination result of the reason for the user's stay. For the vehicles staying at the exit exceeding the limit time, the administrator takes the initiative to contact and assist in providing remote services.

[0052] The beneficial effects of the present invention are as follows:

[0053] 1. By conducting true abnormality, pending abnormality, and rationality determination on the analysis report of the abnormal type list of the vehicles staying in front of the gate, and finally transmitting it to the administrator's device, the present invention can improve the processing efficiency when the administrator finally accepts the user's call for remote service. By conducting a rationality analysis on the possible problems of the current user, it can intelligently confirm the gate abnormality and license plate abnormality in the case of completed payment, so as to improve the processing efficiency of the actual user's remote call service. For the user problems that cannot be intelligently confirmed, by presenting the analysis report of the abnormal type list, it can improve the information interaction efficiency between the administrator and the user, and realize the efficient communication of the parking lot call for remote service;

[0054] 2. By implementing a dynamic grading mechanism for abnormal events, the present invention avoids the need for all abnormal types to be coordinated and solved by contacting the administrator. By presenting the analysis report of the abnormal type list, it avoids the problem of long interaction time caused by the user repeatedly describing the problem and the administrator manually matching the vehicle information. At the same time, for the determination results in specific abnormal situations, it automatically executes the release instruction, reducing the probability of congestion at the parking lot exit;

[0055] 3. By presenting the analyzed report of the marked abnormal type list to the administrator, the present invention provides a systematic basis for the administrator to determine the reasons for user stay, improves the information interaction efficiency between the administrator and the user, and helps to quickly solve the actual problems of the user;

[0056] 4. For abnormal vehicles that stay at the exit beyond the target time and do not contact the administrator, by providing the analyzed report of the current vehicle abnormal type list to the administrator's device, it is convenient for the administrator to understand the possible abnormalities of the current vehicle. At the same time, through pre-data collection and analysis report generation, the information reception efficiency of the administrator is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0058] Figure 1 is a flowchart of a method for visualizing a parking lot to call a remote service according to an embodiment of the present invention;

[0059] Figure 2 is a block diagram of a system for visualizing a parking lot to call a remote service according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] To further illustrate the embodiments, the present invention provides drawings. These drawings are part of the disclosure of the present invention. They are mainly used to illustrate the embodiments and can be used to explain the operating principle of the embodiments in conjunction with the relevant descriptions in the specification. With reference to these contents, those of ordinary skill in the art should be able to understand other possible implementation manners and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are usually used to represent similar components.

[0061] According to an embodiment of the present invention, a method and system for visualizing a parking lot to call a remote service are provided.

[0062] Now, the present invention will be further described in conjunction with the drawings and specific implementation manners;

[0063] Embodiment 1:

[0064] As Figure 1 shown, a method for visualizing a parking lot to call a remote service according to an embodiment of the present invention includes the following steps:

[0065] S1. Based on the response status of the parking lot entrance gate, combined with video data and the changes in parking space status, associate the vehicles entering the parking lot synchronously, make a mark, and record it in the real-time database of the parking lot;

[0066] S11. Collect the parking space status and gate data through the geomagnetic sensors deployed in the parking spaces, as well as the parking lot exit gate and entrance gate, including the parking space occupancy status, parking space occupancy time, vehicle entry time, license plate recognition result, gate lifting status, and gate payment status;

[0067] S12. Obtain the real-time vehicle video data in the parking lot and extract vehicle features through the gate cameras and parking space cameras deployed in the parking lot, including identifying the vehicle color through the color histogram and identifying the vehicle contour through the edge detection algorithm. Combine the parking space occupancy status, vehicle entry time, license plate recognition result, and gate lifting status to associate the vehicles entering the parking lot synchronously, including the following steps:

[0068] According to the vehicle color and contour information collected at the entrance gate, and based on the parking space occupancy status within n minutes after the vehicle entry time, extract the vehicle color and contour information of the occupied parking space within n minutes. Determine the similarity between the vehicle feature data of the parking space and the vehicle feature data of the entrance gate through the Euclidean distance. Based on the similarity threshold, associate the entry time, license plate recognition result, and parking space number of the current vehicle, and generate a mark number. Establish a real-time database of the parking lot through MySQL, use the mark number as the file name, and store relevant information, including the license plate recognition result, parking space number, vehicle entry time, and the results of the edge detection recognition and color histogram reading of the current vehicle.

[0069] It should be noted that n minutes is usually set to 10 minutes, and the extraction time can also be increased or decreased according to the scale and complexity of the parking lot to adapt to the environments of different parking lots. The format of the mark number is "parking lot number + parking space number + time serial number format", for example, "P01_05_2025xxxx1030_003", which represents the vehicle parked in parking space 05 of parking lot P01 at 10:30 on xx / xx / 2025, and 003 represents that the current vehicle is the third vehicle parked in this parking space so far;

[0070] Among them, edge detection algorithms such as Sobel operator and Canny operator are used to detect the vehicle edge by calculating the gradient of pixel points in the video image, so as to identify the vehicle contour. By calculating the color histogram of the vehicle area, the distribution of the number of pixels with different color values in the image is counted. A color template library is established by learning and modeling the histograms of a large number of known color vehicle images. The color histogram of the vehicle in the video to be recognized is compared with that in the template library, and the Bhattacharyya distance is used to measure the similarity, so as to judge the color of the vehicle in the current video. The similarity between the vehicle in the parking space and the vehicle at the entrance is determined based on the Euclidean distance. When both the vehicle contour and color meet the Euclidean distance threshold, it means that the vehicle entering the current parking space is the same as the vehicle at the entrance, and the relevant acquisition data of the vehicle are associated.

[0071] S2. According to the change situation of the parking space state, conduct behavioral logic analysis on the vehicle that causes the change of the parking space state, combine the gate payment state, and generate an analysis report on the current vehicle abnormal type list according to the list of abnormal types of vehicle call remote services in the parking lot, and indicate the abnormal determination result in the analysis report of the abnormal type list, including true abnormality and pending abnormality;

[0072] S21. According to the change result of the occupied state of the parking spaces in the parking lot, conduct behavioral logic analysis on the vehicle that causes the change of the occupied state of the parking space, and combine the occupied time of the parking space to determine the departing vehicle;

[0073] S211. According to the change result of the occupied state of the parking spaces in the parking lot, record the time node when the state of the parking space changes from idle to occupied, and obtain the parking space activation timestamp t start And obtain the license plate recognition result according to the mark number associated with the current parking space number;

[0074] S212. Record the time node when the state of the parking space changes from occupied to idle, and obtain the parking space deactivation timestamp t over , calculate and obtain the usage time Δt of the vehicle in the current parking space = t over -t start , judge whether the vehicle leaving the current parking space is a departing vehicle. When Δt > t θ , determine that the vehicle in the current parking space is a departing vehicle. When Δt ≤ t θ , determine that the current vehicle is a pending vehicle, where t θ is the parking space usage time threshold;

[0075] It should be noted that when determining the departing vehicle, it is mainly to determine whether the current vehicle is a parking space adjustment situation, t θis the time threshold for parking space usage, usually set to 60 minutes, and can also be adjusted according to the situation of different parking lots. By counting the parking time when vehicles in the parking lot drive out of the parking space but do not leave the parking lot within 12 months, select the maximum value plus 10 minutes as the time threshold.

[0076] S213. For the vehicle to be determined, combined with the remaining free parking space status in the parking lot, conduct a secondary determination to determine the actual status of the current vehicle to be determined. The steps include: extracting the relevant data of the parking spaces whose status changes from free to occupied within n / 2 minutes. When there are parking spaces without a marker number generated, extract the vehicle color and contour information of all the parking spaces without a marker number generated, and conduct an Euclidean distance similarity determination with the vehicle edge detection and recognition and color histogram reading results of the current vehicle to be determined in the real-time database of the parking lot. Update the parking space information that meets the similarity determination result to the current vehicle marker number file in the real-time database of the parking lot;

[0077] When there are no parking spaces without a marker number generated, determine the current vehicle as a departing vehicle and make a note in the current vehicle file in the real-time database of the parking lot.

[0078] It should be noted that among them, updating the parking space information that meets the similarity determination result to the current vehicle marker number file in the real-time database of the parking lot means replacing the new parking space number with the old one and synchronously updating and replacing the parking space number in the marker number.

[0079] S22. For the determined departing vehicles, according to the preset list of abnormal types of parking lot vehicle call remote services, combined with the gate payment status and the video data of the exit camera, generate an analysis report of the current vehicle abnormal type list, and note the abnormal relevance and pre-determination results of different abnormal types in the analysis report of the abnormal type list;

[0080] S221. Preset the list of abnormal types of parking lot vehicle call remote services, establish an abnormal set E = {E1, E2, E3, E4}, including E1 payment anomaly, E2 license plate entry anomaly, E3 gate lifting anomaly, E4 other anomalies. The determination feature of each anomaly is that when the gate payment information F pf = 0 and the license plate information F lp = 1, it represents a payment anomaly; when the gate payment information F pf = 1 and the license plate information F lp = 0, it represents a license plate entry anomaly; when the gate payment information F pf = 1 and the license plate information F lp = 1, detect the gate lifting state F up of the exit gate. When F up = 1, it represents other anomalies. When F upWhen it is 0, it represents that the gate lifting of the turnstile is abnormal;

[0081] S222. Extract the data saved in the mark number file of the departing vehicle in the parking lot real-time database, and combine the turnstile payment status and the video data of the exit camera to generate the list data of the abnormal types of the parking lot vehicle call remote service for the vehicle staying at the current exit turnstile:

[0082] Identify the license plate recognition result of the vehicle staying in front of the current exit turnstile through the exit turnstile, and traverse the license plate recognition results in the mark number file in the parking lot real-time database. When there is a consistent license plate, determine that the license plate information F of the vehicle staying in front of the current exit turnstile lp = 1. When there is no consistent license plate, determine that the license plate information F of the vehicle staying in front of the current exit turnstile lp = 0;

[0083] For the case where the license plate information F lp = 1, obtain the turnstile payment information of the current license plate vehicle through the exit turnstile. The payment status is F pf = 1 for paid, and F pf = 0 for unpaid;

[0084] For the case where the license plate information F lp = 0, retrieve all the actually received payment amounts M up to now 实收 and the payment amount M of the departing vehicle 出库 , calculate the amount anomaly value ΔM = M 实收 - M 出库 . When ΔM = 0, it is determined that the payment has been made and F lp = 1. When ΔM < 0, it is determined that the payment has not been made and F pf = 0;

[0085] It should be noted that by comparing all the actually received payment amounts M 实收 with the payment amount M of the departing vehicle 出库 , it is possible to identify the situation where the license plate numbers of the same vehicle are misrecognized at the entrance turnstile and the exit turnstile. When ΔM = 0, it means that the current vehicle owner has completed the payment according to the misrecognized license plate number at the entrance turnstile, which is convenient for subsequent determination of the vehicle status before calling the remote service.

[0086] When the exit turnstile completes the lifting of the rod, record F up = 1. When the exit turnstile fails to complete the lifting of the rod, record F up = 0;

[0087] S223. For the abnormal entry of E2 license plates and the abnormal lifting of the E3 gate, obtain the video data of the vehicle staying at the current exit gate through the video data of the exit camera, extract the features of the current staying vehicle through the color histogram and edge detection algorithms, and perform Euclidean distance similarity determination by combining the vehicle features in the current departing vehicle mark number file. When there is a matching mark number file, it is determined that the current abnormal entry of the E2 license plate and the abnormal lifting of the E3 gate are true abnormalities. When there is no matching mark number file, the abnormality is determined to be a pending abnormality;

[0088] S224. Based on the F lp 、F pf 、F up of the current staying vehicle, generate an analysis report on the list of abnormal types of the current vehicle. The analysis report includes the abnormal type E x of the current staying vehicle, where the value range of x is 1-4. When x = 2 or x = 3, the abnormality is synchronously marked as a true abnormality or a pending abnormality.

[0089] It should be noted that when the features of the staying vehicle are similar to those of the vehicle in the departing vehicle mark number file in terms of Euclidean distance determination, it indicates that there is a high possibility of abnormal license plate entry and abnormal gate lifting. Marking it as a true abnormality is convenient for subsequent direct customer service.

[0090] Embodiment 2:

[0091] S3. For the remote service content called by the user of the current staying vehicle, combine the user call content with the keyword libraries of different abnormal types to determine the consistency between the current user call content and the abnormal types in the analysis report on the list of abnormal types, and mark the rationality. Based on the determination results of different abnormal types, output actual solutions, including gate release and manual transfer, and at the same time present the analysis report on the list of abnormal types of the current vehicle after marking to the administrator device, so as to systematically provide the basis for the determination result of the user staying reason. For the exit staying vehicles exceeding the restricted time, the administrator takes the initiative to contact to assist in remote service;

[0092] S31. Transmit the analysis report on the list of abnormal types of the current vehicle to the administrator device, record the remote service call content of the current vehicle, extract the keywords in the remote service call content of the current vehicle through natural language processing technology, and determine the abnormality rationality by combining the abnormal types and keywords in the analysis report on the list of abnormal types of the current vehicle;

[0093] S311. Preset the keyword set k corresponding to each abnormal type E where k represents the abnormal type and N represents the keyword number, and assign weights to each keyword The content of the current vehicle's remote service call is converted from speech to text through Speech-to-Text. The text is segmented by the forward maximum matching method, stop words are removed, and core keywords are retained, and the segmented set G = {g1, g2,..., g M} is obtained. Based on the preset synonym table, the words in the segmented set are replaced with standard keywords;

[0094] It should be noted that the keywords for E1 payment exception are "payment failure", "abnormal payment amount", "missing payment record", etc., the keywords for E2 license plate entry exception are "license plate", "recognition error", "not entered", "information mismatch", etc., and the keywords for E3 barrier lifting exception are "lifting the barrier", "payment", "paying", etc. The synonym table needs to be set based on the actual application environment of the parking lot. The influencing factors include dialects and common sayings. For example, "can't pay" in text recognition segmentation is replaced with the standard keyword "payment failure".

[0095] S312. For each exception type E k , count the number and weight of keywords in the user's text that match its keywords, convert the matching weight into a percentage form, and select the exception type with the highest percentage as the target exception type in the content of the current vehicle's remote service call;

[0096] It should be noted that converting the matching weight into a percentage form can avoid the deviation caused by different numbers of keywords and ensure the accuracy of the matching. The percentage conversion formula is:

[0097]

[0098] where U(·) is an indicator function, which is 1 when the user's keyword belongs to the keyword library and 0 otherwise. Exception type E k The preset weight of the I-th keyword, represents the sum of the weights of all keywords of exception type E k , which is used to eliminate the deviation caused by the difference in the number of keywords, and the result is expressed in percentage form. The larger the value, the higher the matching degree between the user's text and exception type E k .

[0099] S313. Count the exception types in the exception type list analysis report of the current vehicle. When the exception type in the exception type list analysis report of the current vehicle is the same as the target exception type, it means that the current vehicle's exception is reasonable. When the exception type is inconsistent with the target exception type, it means that the current vehicle's exception is unreasonable, and it is recorded in the exception type list analysis report of the current vehicle to obtain the rationality analysis report of the exception type list.

[0100] S32. Based on the abnormal rationality determination result, take targeted measures for the user's remote service call for service;

[0101] S321. When the user's abnormal type is E2 or E3 and both the true abnormality and abnormal rationality are determined, directly perform the operation of releasing the turnstile;

[0102] S322. When the user's abnormal type does not meet the requirements of S321, output the rationality analysis report of the abnormal type list of the current vehicle to the administrator's device and perform the manual transfer operation.

[0103] It should be noted that by analyzing the report of the abnormal type list of the vehicle staying in front of the turnstile to determine the true abnormality, pending abnormality and rationality, and finally transmitting it to the administrator's device, the processing efficiency when the administrator finally accepts the user's call for remote service can be improved. By analyzing the rationality of the possible problems of the current user, intelligent confirmation is carried out for the turnstile abnormality and license plate abnormality in the case of completed payment, so as to improve the processing efficiency of the actual user's remote call service. For the user problems that cannot be intelligently confirmed, by presenting the analysis report of the abnormal type list, the information interaction efficiency between the administrator and the user is improved.

[0104] S3221. Identify the abnormal vehicle at the exit through the camera at the exit turnstile based on the edge detection algorithm, record the start time stamp of the abnormal vehicle occupying the exit, and obtain the cumulative duration t of the abnormal vehicle occupying the exit 累计 , when t 累计 exceeds 5 minutes and the administrator's device has not received the user's request for remote service call, actively transmit the analysis report of the current vehicle's abnormal type list to the administrator's device;

[0105] S3221. Connect the communication device deployed at the exit turnstile through the communication device at the administrator's device to actively provide services for the current abnormal vehicle.

[0106] It should be noted that by providing the analysis report of the current vehicle's abnormal type list to the administrator's device, it is convenient for the administrator to understand the possible abnormalities of the current vehicle. At the same time, through the pre-data collection and analysis report generation, the administrator's information reception efficiency is improved. The communication device includes a voice intercom, a telephone, etc.

[0107] S4. According to the sensor information of the exit turnstile, for the vehicles that have left the parking lot, make a note number file in the real-time database of the parking lot.

[0108] Embodiment 3:

[0109] Such as Figure 2As shown in the figure, a visual parking lot call remote service system includes a vehicle data collection module, a vehicle behavior analysis module, an exception list generation module, a call content determination module, and a real-time operation module;

[0110] The vehicle data collection module collects the parking space status and gate machine data through the geomagnetic sensors deployed in the parking spaces, as well as the exit gate machine and the entrance gate machine of the parking lot, and collects the vehicle video picture data through the cameras deployed inside the parking lot;

[0111] The vehicle behavior analysis module associates the vehicles entering the parking lot synchronously based on the response status of the entrance gate machine of the parking lot, combines the video data and the change situation of the parking space status, makes a mark, and records it in the real-time database of the parking lot;

[0112] The exception list generation module conducts a behavioral logic analysis on the vehicles that cause the change of the parking space status according to the change situation of the parking space status, combines the gate machine payment status, and generates an analysis report of the current vehicle exception type list according to the exception type list of the parking lot vehicle call remote service, and indicates the exception determination result in the analysis report of the exception type list, including true exceptions and pending exceptions;

[0113] The call content determination module conducts a consistency determination on the current user's call remote service content for the vehicle staying currently, combines the user's call content with the keyword libraries of different exception types, and determines the rationality;

[0114] The real-time operation module outputs actual solutions based on the determination results of different exception types, including gate machine release and manual transfer, and at the same time presents the analysis report of the current vehicle exception type list after marking to the administrator device, so as to systematically provide the basis for the determination result of the user's staying reason, and actively contact the administrator for the vehicles staying at the exit exceeding the restricted time to assist in remote service.

[0115] In summary, the present invention can improve the processing efficiency when the administrator finally accepts the user's call for remote service by conducting true exception, pending exception, and rationality determination on the analysis report of the exception type list of the vehicle staying in front of the gate machine, and finally delivering it to the administrator device. By conducting a rationality analysis on the possible problems of the current user, it can intelligently confirm the gate machine exception and license plate exception in the case of completed payment to improve the processing efficiency of the actual user's remote call service. For the user problems that cannot be intelligently confirmed, by presenting the analysis report of the exception type list, it can improve the information interaction efficiency between the administrator and the user, and realize the efficient communication of the parking lot call remote service.

[0116] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for visualizing a parking lot call remote service, characterized in that, The method includes the following steps: S1. Based on the response status of the parking lot entrance gate, combined with video data and the change of parking space status, associate the vehicles entering the parking lot synchronously, make a mark, and record it in the real-time database of the parking lot; S2. According to the change of parking space status, conduct behavioral logic analysis on the vehicles that cause the change of parking space status, combined with the gate payment status, and generate an analysis report on the current vehicle exception type list according to the list of abnormal types of remote services called by vehicles in the parking lot. Indicate the abnormal determination result in the analysis report of the exception type list, including true exception and pending exception; S3. For the content of the remote service called by the user of the currently parked vehicle, combined with the user call content and the keyword library of different exception types, determine the consistency between the current user call content and the exception types in the analysis report of the exception type list, and mark the rationality. Based on the determination results of different exception types, output the actual solutions, including gate release and manual transfer. At the same time, present the analysis report of the current vehicle exception type list after marking to the administrator device to systematically provide the basis for the determination result of the user's staying reason. Actively contact the administrator for the vehicles staying at the exit exceeding the restricted time to assist in remote services; S4. According to the sensor information of the exit gate, for the vehicles that have left the parking lot, make a note of the mark number file in the real-time database of the parking lot.

2. The visualization parking lot call remote service method according to claim 1, characterized in that The S1 includes the following steps: S11. Collect the parking space status and gate data through the geomagnetic sensors deployed in the parking spaces, the parking lot exit gate, and the entrance gate, including the parking space occupancy status, parking space occupancy time, vehicle entry time, license plate recognition result, gate lifting status, and gate payment status; S12. Obtain the real-time vehicle video data in the parking lot through the gate cameras and parking space cameras deployed in the parking lot and extract vehicle features, including identifying the vehicle color through the color histogram and identifying the vehicle contour through the edge detection algorithm. Combine the parking space occupancy status, vehicle entry time, license plate recognition result, and gate lifting status to associate the vehicles entering the parking lot synchronously, including the following steps: According to the vehicle color and contour information collected at the entrance gate, according to the parking space occupancy status within n minutes after the vehicle entry time, extract the vehicle color and contour information of the occupied parking space when the parking space is occupied within n minutes. Determine the similarity between the vehicle feature data of the parking space and the vehicle feature data of the entrance gate through the Euclidean distance. Based on the similarity threshold, associate the entry time, license plate recognition result, and parking space number of the current vehicle, and generate a mark number. Establish a real-time database of the parking lot through MySQL, use the mark number as the file name, and store relevant information, including the license plate recognition result, parking space number, vehicle entry time, and the results of edge detection recognition and color histogram reading of the current vehicle.

3. The visual parking lot call remote service method according to claim 2, characterized in that, The S2 includes the following steps: S21. According to the change result of the parking space occupancy status in the parking lot, conduct behavioral logic analysis on the vehicles that cause the change of the parking space occupancy status, combined with the parking space occupancy time, to determine the departing vehicles; S22. For the determined departing vehicle, based on the preset list of abnormal types of parking lot vehicle call remote services, combined with the gate payment status and the video data of the exit camera, generate an analysis report of the current vehicle's abnormal type list, and indicate the abnormal relevance and preliminary determination results of different abnormal types in the analysis report of the abnormal type list.

4. The visual parking lot call remote service method according to claim 3, characterized in that, The S21 includes the following steps: S211. Record the time node when the parking space status changes from idle to occupied according to the change result of the occupancy status of the parking spaces in the parking lot, and obtain the parking space activation timestamp t start And obtain the license plate recognition result according to the notation number associated with the current parking space number; S212. Record the time node when the parking space status changes from occupied to free to obtain the parking space deactivation timestamp t over , and calculate the vehicle usage time Δt of the current parking space as Δt = t over - t start . Determine whether the vehicle leaving the current parking space is a departing vehicle. When Δt > t θ , it is determined that the vehicle in the current parking space is a departing vehicle. When Δt ≤ t θ , it is determined that the current vehicle is a pending vehicle, where t θ is the parking space usage time threshold; S213. For the undetermined vehicle, make a secondary determination in combination with the remaining free parking space status in the parking lot to determine the actual status of the current undetermined vehicle. The steps include: extracting the parking space - related data where the parking space status changes from free to occupied within n / 2 minutes. When there are parking spaces without a marker number generated, extract the color and contour information of the vehicles occupying these parking spaces, and perform Euclidean distance similarity determination with the vehicle edge detection recognition and color histogram reading results of the current undetermined vehicle in the parking lot real - time database. Update the parking space information that meets the similarity determination result to the marker number file of the current undetermined vehicle in the parking lot real - time database; When there are no parking spaces without a marker number generated, determine the current vehicle as a departing vehicle and make a note in the current vehicle file in the parking lot real - time database.

5. A method for visualizing a parking lot call remote service according to claim 3, characterized in that The S22 includes the following steps: S221. Preset a list of abnormal types for vehicle calls to remote services in a parking lot, and establish an abnormal set E = {E1, E2, E3, E4}, including E1 payment exception, E2 license plate entry exception, E3 gate lifting exception, and E4 other exceptions. The determination feature for each exception is that when the gate payment information F pf = 0 and the license plate information F lp = 1, it represents a payment exception; when the gate payment information F pf = 1 and the license plate information F lp = 0, it represents a license plate entry exception; when the gate payment information F pf = 1 and the license plate information F lp = 1, detect the gate lifting status F up of the exit gate. When F up = 1, it represents other exceptions. When F up = 0, it represents a gate lifting exception; S222. Extract the saved data in the marker number file of the departing vehicle in the parking lot real - time database, and generate the data of the abnormal type list of parking lot vehicle call remote services for the vehicle staying at the current exit gate in combination with the gate payment status and the video data of the exit camera: Identify the license plate recognition result of the vehicle staying in front of the current exit gate through the exit gate, and traverse the license plate recognition results in the mark number file in the real-time database of the parking lot. When there is a consistent license plate, determine that the license plate information F of the vehicle staying in front of the current exit gate lp = 1. When there is no consistent license plate, determine that the license plate information F of the vehicle staying in front of the current exit gate lp = 0; For license plate information F lp In the case of lp = 1, obtain the gate payment information of the vehicle with the current license plate through the exit gate. The payment status is F pf pf = 1 indicates that the payment has been made, and pf = 0 indicates that the payment has not been made pf = 0; For license plate information F lp In the case of = 0, retrieve all the actually received payment amounts M up to now 实收 and the payment amount M of the departing vehicles 出库 , calculate the amount anomaly value ΔM = M 实收 -M 出库 , when ΔM = 0, it is determined that the payment has been made F lp = 1, when ΔM < 0, it is determined that the payment has not been made F pf = 0; When the exit turnstile completes the raising of the rod, record F up = 1, when the exit turnstile does not complete the raising of the rod, record F up = 0; S223. For the abnormal E2 license plate entry and abnormal E3 gate lifting, obtain the video data of the vehicle staying at the current exit gate through the video data of the exit camera, extract the features of the current staying vehicle through the color histogram and edge detection algorithm, and perform Euclidean distance similarity determination in combination with the vehicle features in the marker number file of the current departing vehicle. When there is a matching marker number file, determine that the current abnormal E2 license plate entry and abnormal E3 gate lifting are true abnormalities. When there is no matching marker number file, determine the abnormality as a pending abnormality; S224. Based on the F of the currently parked vehicle lp F pf F up Generate an analysis report on the list of abnormal types of the current vehicle. The analysis report includes the abnormal type E of the currently parked vehicle x where the value range of x is 1 - 4. When x = 2 or x = 3, synchronously mark the abnormality as a true abnormality or a pending abnormality.

6. The visualized parking lot call remote service method according to claim 5, wherein The S3 includes the following sub - steps: S31. Transmit the analysis report of the abnormal type list of the current vehicle to the administrator device, record the content of the current vehicle's remote service call, extract the keywords in the content of the current vehicle's remote service call through natural language processing technology, and determine the abnormality rationality in combination with the abnormal types and keywords in the analysis report of the abnormal type list of the current vehicle; S32. Based on the abnormality rationality determination result, take targeted measures to serve the user's remote service call.

7. A method for visualizing a parking lot call remote service according to claim 6, characterized in that, The S31 includes the following steps: S311. Preset each exception type E k The corresponding keyword set where k represents the exception type, N represents the keyword number, and weights are assigned to each keyword Perform speech-to-text conversion on the current vehicle remote service call content to obtain text, perform word segmentation on the text by the forward maximum matching method, remove stop words and retain core keywords, and output the word segmentation set G = {g1, g2,..., g M}, and based on the preset synonym table, replace the words in the word segmentation set with standard keywords; S312. For each exception type E k , count the number and weight of keywords in the user text that match it, convert the matching weight into a percentage form, and select the exception type with the highest percentage as the target exception type in the current vehicle remote service call content; S313. Count the abnormal types in the analysis report of the abnormal type list of the current vehicle. When the abnormal types in the analysis report of the abnormal type list of the current vehicle are the same as the target abnormal types, it means that the current vehicle's abnormality is reasonable. When the abnormal types are different from the target abnormal types, it means that the current vehicle's abnormality is unreasonable, and record it in the analysis report of the abnormal type list of the current vehicle to obtain the analysis report of the abnormality type list rationality.

8. A method for visualizing a parking lot call remote service according to claim 7, characterized in that, The S32 includes the following steps: S321. When the user exception type is E2 or E3 and both the true exception and the exception rationality determination are satisfied, the turnstile release operation is directly performed. S322. When the user exception type does not meet the requirements of S321, an analysis report on the rationality of the exception type list of the current vehicle is output to the administrator device, and a manual transfer operation is performed.

9. A method for visualizing a parking lot call remote service according to claim 8, characterized in that, The S322 includes the following steps: S3221. Use the camera at the exit turnstile to identify abnormal vehicles based on the edge detection algorithm, record the starting timestamp of the abnormal vehicle occupying the exit, and obtain the cumulative duration t of the abnormal vehicle occupying the exit. 累计 When 累计 t exceeds 5 minutes and the administrator device has not received a user call for remote service request, actively transmit the analysis report of the current vehicle abnormal type list to the administrator device. S3221. Connect to the communication device deployed at the exit turnstile through the communication device at the administrator device to actively provide services for the current abnormal vehicle.

10. A visual parking lot call remote service system, characterized in that, This system adopts a visual parking lot call remote service method as described in any one of claims 1-9, including a vehicle data collection module, a vehicle behavior analysis module, an exception list generation module, a call content determination module, and a real-time operation module. The vehicle data collection module collects the parking space status and turnstile data through the geomagnetic sensors deployed in the parking spaces, the exit turnstiles, and the entrance turnstiles of the parking lot, and collects the vehicle video image data through the cameras deployed inside the parking lot. The vehicle behavior analysis module, based on the response status of the entrance turnstile of the parking lot, combines the video data and the change of the parking space status, correlates the vehicles entering the parking lot synchronously, makes a mark, and records it in the real-time database of the parking lot. The exception list generation module analyzes the behavior logic of the vehicles that cause the change of the parking space status according to the change of the parking space status, combines the turnstile payment status, and generates an analysis report on the exception type list of the current vehicle according to the exception type list of the parking lot vehicle call remote service, and indicates the exception determination result in the analysis report on the exception type list, including true exception and pending exception. The call content determination module determines the consistency between the current user call content and the exception type in the analysis report on the exception type list by combining the user call content with the keyword library of different exception types for the user call remote service content of the currently parked vehicle, and marks the rationality. The real-time operation module outputs actual solutions based on the determination results of different exception types, including turnstile release and manual transfer, and at the same time presents the analysis report on the exception type list of the current vehicle after marking to the administrator device to systematically provide the basis for the determination result of the user's stay reason, actively contacts the administrator for the exit stay vehicle that exceeds the limit time, and assists in providing remote services.

Images